Direct Democracy and linguistic minorities in Switzerland and South Tyrol – A comparison
Thomas Benedikter
Abstract
Do procedures of direct democracy involve fundamental dangers with regard to social, ethno-linguistic and religious minorities? Is there even a conflict between popular decision-making by initiative and referendum and the requirement of interethnic consociational democratic procedures? The article focuses on the existing tension and supposed contradictions between procedures of direct democracy (initiative and referendum) and the protection of the rights and interests of minorities in multilingual areas. Drawing upon some recent examples of such kind in four Swiss bi- or multilingual cantons and in South Tyrol the author compares these experiences and discusses the crucial issue as to whether such a conflict can be solved and what special precautions should be established when it comes to applying direct democracy in multilingual areas. Finally he presents a new approach to how civil rights to direct political participation, enshrined in Italy's constitutions and South Tyrol's statute of autonomy, might be brought into concordance with the need to respect the equality of rights of linguistic groups within a multilingual polity, an issue which is currently being debated in South Tyrol's politics.
Introductory remarks
Do procedures of direct democracy encompass fundamental dangers with regard to social, ethno-linguistic and religious minorities? Is there even a conflict between popular decision-making by initiative and referendum and the requirement to protect linguistic minorities and consociational democratic procedures in multilingual regions? The existing tension between democracy and the rule of law (including minority rights), and between democratic majority rule and the protection of minorities is widely known. Critics of direct democracy assert that majorities of the population neither behave in a rational nor in an altruistic manner. Thus, it is alleged that in referendums concerning minority issues voters would cast their votes in a less responsible manner than elected politicians. The initiative to deny Muslims in Switzerland the right to build new minarets, recently accepted by the Swiss electorate, has fuelled such allegations, overlooking the fact that in previous referendums Swiss voters have mostly voted in a minority- and foreigner-friendly way.[1] Kirchgässner (2009, 5) argues: “There is no empirical evidence that Swiss voters, when called to vote on initiatives and referendums, systematically discriminate against minorities; it cannot be stated unequivocally whether civil rights are more effectively protected in a system with referendum rights or one without such rights”.[2]
Another study (Frey et al. 1998, 1343-1348, at 1343) concluded: “Out of 64 popular votes on minority rights in Switzerland between 1970 and 1996, only 19 (30%) brought about a restriction of minority rights. In more than two-thirds of all cases Swiss voters strengthened minority rights, instead of limiting them. Therefore direct democracy protects civil rights”.[3]
In respect of Germany, the advocacy NGO Mehr Demokratie e.V. came to a similar conclusion: “Only two out of 131 requests for popular initiatives at the level of the Länder encompassed provisions hostile to minorities. Both were blocked by the Constitutional Court and did not even reach the ballot vote”.[4]
Procedures of direct democracy are designed to express the real political preferences of the citizens - which often differ from those of the elected representatives. Discrimination against minorities happens in both systems: in systems vested with instruments of direct democracy and in purely representative systems (Verhulst et al. 2007, 70). Majority decisions in referendums may address fundamental rights, as observed in the cases of the ban on the building of further minarets in Switzerland and the proposal to ban gay marriages in the USA. But when in a referendum a majority of the electorate accepts a proposal which is at the expense of a minority group, be it a social group, or one defined by sexual orientation, religion or language, this may certainly be to the disadvantage of the concerned minority; but it cannot automatically be asserted that fundamental rights or human rights have been denied or violated.
On the one hand some scholars and some media commentators tend to overestimate the minority-friendliness and accountability of politicians, which is supposed to obtain in purely representative systems. At the parliamentary level minority rights and interests are very often limited or regulated in a way which has not been agreed with the concerned minorities. As an example we may think of the rights of foreign migrant workers and asylum seekers in many European countries.
On the other hand, in Europe minorities are not completely exposed to arbitrary decisions of a popular majority, as both the European Convention on Human Rights and the EU Charter of Fundamental Rights, along with the national constitutions, protect minority rights quite rigorously. Moreover, with regard to linguistic minorities, the jurisdiction of the European Court of Human Rights, based on the European Convention of Human Rights, safeguards their rights. Kirchgässner (2009, 19) states: “It is quite well proven empirically that direct-democratic rights urge politicians to follow popular preferences to a much larger extent. Thus, the recognition of minority rights mainly depends on those preferences. If a clear majority of the population favours the restriction (or extension) of such rights, this will be mirrored in policies: in a purely representative system as well as in a system with direct democracy”.
Considering participatory procedures and the rights of minorities from this perspective two major issues can be raised. First, what are the boundaries a society wishes to set around minority and human rights as a sort of “untouchable core-area” of rights which may not be challenged even by the sovereign citizens? In other terms: what rights are to be accorded to minorities regardless of democratic majorities resulting from the ballot?
Second, what kind of institutional arrangements should be established for the application of instruments of direct democracy in order to prevent abuse and discrimination?
With respect to the rights of ethno-linguistic minorities, what rights of such minorities should be excluded from referendums and how should direct democracy in the hands of all citizens be regulated in multilingual regions, in order to prevent the violation of minority rights and the alienation of the groups from each other? In this essay some examples of popular votes from bi- and multilingual regions in the Alps – four Swiss cantons (Wallis, Fribourg, Graubünden and Berne) and the autonomous province of South Tyrol in Italy – will be considered and compared, before some conclusions are drawn.[5]
Swiss experiences with direct democracy in bi- and multilingual cantons
In Switzerland, the multi-ethnic cradle of direct democracy, the interplay between popular votes and the relationship of minority and majority groups can be analysed with reference to the four official language groups. First, we can examine whether there have been significant discrepancies between the linguistic territories in the voters' behaviour in popular referendums at the federal level. Second, we will see what happened in multilingual cantons when linguistically sensitive issues were put to a popular vote.
According to an evaluation of the results of referendums at the federal level from 1875 to 1994 carried out by Kriesi et al. (1996), in 29 out of 430 referendums there was a difference of more than 25% in the rate of approval of initiatives among German and French speakers. But only in 15 cases (3.5% of all votes) were the French speakers outvoted. In 39 cases a significant difference between German and Italian speakers is reported.[6] In the period 1983-1994, a difference of more than 20% between French and German speakers is reported on 10 out of 116 occasions. But in only four cases were the French speakers outvoted, whereas the Italian speakers were outvoted in six cases, in other terms: in six cases the outcome of the polls in the canton of Ticino was markedly different from that in polls at the federal level. In absolute terms, the number of cases of “outvoting” of Italian and French speaking Swiss has increased since the 1970s, but the significance of such discrepancies in electoral behaviour among the Swiss language groups has been exaggerated by the French language media.
Nenad Stojanovic (2006) carried out a study into the connection between the alleged growing differences between Switzerland's two major linguistic communities when using the instruments of direct democracy. For this purpose he evaluated the results of the referendums in four multilingual cantons, as the Swiss practice of direct democracy is of even greater significance at the municipal and cantonal levels, as compared to the federal level. 22 out of 26 Swiss cantons are monolingual, four are bilingual or multilingual.[7]
Table 1: Ethno-linguistic composition of bi- and multilingual cantons of Switzerland expressed as a %; (data from the census of 2001)
Language groups / German / French / Italian / Rhaetoromanian (or Ladin in Italy)Wallis / 28.4 / 62.8 / 2.2 / -
Graubünden / 68.3 / 0.5 / 10.2 / 14.5
Freiburg / 29.2 / 63.2 / 1.3 / 0.1
Bern / 84.0 / 7.6 / 2.0 / 0.1
Switzerland / 63.7 / 20.4 / 6.5 / 0.5
Comparison: South Tyrol (census 2001) / 69.1 / - / 26.5 / 4.4
Source: [8]
Map 1: Switzerland's linguistic composition
In all four bilingual cantons the three main instruments of direct democracy are established: the obligatory referendum whenever the cantonal constitution is to be amended; the optional referendum (on cantonal laws); the legislative and the constitutional initiative. The numbers of signatures required for requesting such procedures in all four cantons are quite low. What has been the impact of direct democracy in those four cantons? Stojanovic focused on two issues in his research: first, whether the respective linguistic minority has been outvoted in a referendum; second, whether there have been popular votes on linguistically sensitive issues and what their outcomes were (Stojanovic 2006). In the present essay I will focus on some examples of the latter question.
A quite emblematic case in this regard was the so-called “language war” in Fribourg/Freiburg in 2000. Under a new cantonal law, 10-15% of the lessons in public schools were required to be held in the second language of the canton, German or French respectively. Some citizens initiated a facultative referendum against this law. On 24 September 2000 they succeeded in abolishing the law by a popular vote against the majority of the canton's political elite. Whereas 70% of the voters in the German-speaking districts had accepted the law, in the French- speaking districts of Fribourg only 40% of the voters had agreed with it. The referendum campaign mobilised both ethno-linguistic groups. Among both groups fears and resistance were fostered against alleged attempts at “Germanification” or “Frenchification”. With ethno-linguistic arguments dominating the referendum campaign, there were reports of an open confrontation between the two language groups such as had not occurred for a long time. In the end, the law was abolished [or ‘struck down’ or ‘repealed’].
In the trilingual canton of Graubünden/Grischun/Grigioni a federal referendum became a test for the relationship between the language groups. In 1996, an amendment to the Swiss constitution was aimed at allowing federal subsidies to be made available for furthering the Italian and Rhaeto-Romanian languages in Ticino and Graubünden. In the federal referendum the proposal was accepted by a majority of 76% of the votes, though in Graubünden the overall approval rate was only 68%. In those Graubünden districts with a clear German majority d the proposal was actually rejected. The canton's media and the Lia Rumantscha, an umbrella organization for Rhaeto-Romanian speakers, blamed the opposition by the German-speaking parts of Graubünden for the weaker cantonal support for the enhancement of the Rhaeto-Romanian language.
The most interesting case was the referendum on the reform of cantonal electoral law in the canton of Wallis/Valais on 26 September 2005. The main issue of this popular initiative was the appropriate level of representation of German-speaking Walliser in the cantonal government, certainly an “ethnically” very sensitive issue. The initiative aimed to introduce a strictly proportional system for electing members to the 5-member cantonal government. This would mean that the number of representatives of the German speakers of Wallis would be reduced from two to only one. Most people in the German speaking part of the canton, the upper Wallis, feared the possible loss of one of their two seats and were inclined to reject the initiative. In the run-up to the referendum, the cantonal government warned against the perceived threat to the unity of the canton. In practice, a majority of Walliser (54%) rejected the initiative proposal in the referendum, though 51% of voters from the French-speaking lower Wallis voted in favour of it. Decisive for the final result was the clear opposition (69% against) of the German-speaking upper Wallis. The linguistic minority won, but what would have happened if the German Upper Wallis had been defeated?
In the canton of Berne just 7.6% of the resident population (concentrated in the Jura region of the canton) speak French as their main language. In April 2006, a reform of the cantonal electoral system was envisaged. In the proposal, the number of constituencies was to be reduced from 27 to 8, and the number of seats in the council(cantonal parliament) cut from 200 to 160. However, in order to ensure a proper representation of the Jura district, it was proposed that 12 seats should be assigned to this area. In the newly constituted district of Biel-Seeland the French speakers should have a fixed share of seats according to their proportion of the total population (about 5%).
Being an amendment to the constitution, the reform had to pass the obligatory referendum, driving the politicians to seek the utmost consent of the electorate. However, objections were raised that the proposed regulation could disrupt the linguistic consensus within the canton, as the voters were being asked to see themselves as members of a linguistic minority, although many of them had already developed a multilingual self-perception. Many French speakers did not wish to be labeled a “protected linguistic minority”. At the cantonal level the constitutional initiative was eventually accepted with a large majority of 84% of YES votes, whereas in the district of Biel the consensus position won the support of only 79% of the voters.
To sum up, only very few cases are reported from these four cantons of referenda or initiatives on linguistically sensitive issues bringing about a contrasting response along linguistic lines. The cantonal constitutions do not exclude such issues from being submitted to popular vote, but such issues are rarely raised. It is part of the 140-yearsold practice of direct democracy in Switzerland to ensure the utmost sovereignty to the citizens on the full range of political issues. The application of popular rights in democratic decision-making did not cause any major friction between the Swiss language groups within the bi- or multilingual cantons (the wider story of the formation of the canton of Jura in the 1970s should be considered an exception); the differences in the results of popular votes at the federal level between the single linguistic territories (Romandie, Deutsch-Schweiz, Ticino) has not so far been used as a pretext for limiting initiative and referendum rights for the citizens.
Experiences in South Tyrol
As a fifth example of referenda held in a multilingual region of the Alps we may consider two cases in South Tyrol (the Autonomous Province of Bozen/Alto Adige) which, although being autonomous, does not have such a large historical tradition of direct democracy as neighbouring Switzerland. The first example is that of a municipal referendum in Bozen, the capital city of South Tyrol (102,000 inhabitants), on a proposed re-naming of a square in autumn 2002; the second is the first provincial referendum on five proposals on 25 October 2009. It should be mentioned that South Tyrol's statute of autonomy is based on the recognition of three official language groups – the German, Italian and Ladin groups. Around 70% of the population speak German, 25% Italian, and almost 5% Ladin. But for historical reasons the territorial pattern of settlement of these groups is quite uneven, showing a tendency to territorial homogeneous clusters. Italian speakers are in a majority in only five out of the 116 municipalities, while the Ladin speakers are concentrated in eight municipalities in the two valleys of Gardena and Badia. 90% of the Italian speakers of South Tyrol live in the major cities. 103 municipalities have a majority of German speakers.
Map 2: The territorial distribution of linguistic groups in South Tyrol (2001)
Source: Autonomous Province of South Tyrol; Rep. 8 – Provincial Statistics Institute, Census 2001.
Bozen, autumn 2002: a municipal referendum to change a place name
In Bozen, the capital of the autonomous South Tyrol, a monument is causing permanent annoyance among the German speaking population, triggering tension and resentment between the two major linguistic groups. The so-called “Victory Monument” was erected in 1928 by order of Mussolini to celebrate the occupation and annexation of South Tyrol to Italy. The surrounding square was given the name “Victory Square” by the fascist administration of the time. Unlike the former fascist monuments in the remaining part of Italy, the triumphal arch in Bozen not only has not been removed, but it has been repeatedly restored by successive Italian governments. It bears all the core symbols of Italian fascism and some writings which mirror the colonial mind-set of its builders.[9] It was only very recently that some information boards had been placed beneath the monument to explain its historical significance was to the local population and to visitors to the city. After many years of argument, on 15 November 2001 the city’s mayor, backed by the town council, decided to rename the square the “Peace Square”, instead of “Victory Square”. On Bozen's city council there was a majority of Italian centre-left parties in coalition with the German SVP and the “inter-ethnic” Green Party, voted by electors from all ethno-linguistic groups. German speakers make up some 26% of the population of Bozen , whereas Italian speakers account for 73% (census 2001).
The mayor’s action provoked a strong reaction from nationalist Italian groups and raised some popular protest. The two right-wing parties, Forza Italia and Unitalia, which are backed by the majority of Italian voters in Bozen, demanded a municipal referendum aimed at compelling the administration to withdraw the new name of the square. In the referendum held on 6 October 2002, 62% of the voters voted to restore the original name “Victory Square”. This referendum made it clear that the fascist symbols and this monument as a cornerstone of the “Italianness” of the town are still held in high esteem by the majority of the Italian speakers of Bozen. Moreover, as the Italian right-wing forces were able to mobilize the Italian electorate, the linguistic minorities had no chance of countering the power of the symbolism within the framework of a society still polarized along ethno-linguistic lines (Kurth et al. 2006).